PhD theses that contributed to GeoDiverCity programme

PhD Clémentine Cottineau

Cottineau C., 2014, The evolution of cities in the post-Soviet Space. Observation and modellings, University Paris I.

Abstract. The Russian and Soviet urbanisation process happened late and fast, compared to other territories. Many new towns and cities were created by the Soviet regime that officially promoted discourses about the socialist function of the city, rational organisation of space and a planned management of the economy. These urban particularities and the multiple demographic and political events of the 20th century have made cities in the post-Soviet space an interesting case and raised question regarding its ruptures and comparability. This dissertation thesis aims to show that the concept of system of cities and generic methods in urban geography (especially models) are useful in the study of urban evolution over the long term in this space, to eventually better understand past trends and predict future ones. We confronted several statistical models with the observed urban dynamics and concluded that the macro-geographical structure of cities in the post-Soviet space was comparable to that of other systems of cities (hierarchy of sizes, spacing, functional differentiation). We also observed specific trajectories related to the size of the territory, natural resources, the recent demographic shrinkage and the effect of particular political decisions. This knowledge about observed evolutions has been included in an incremental approach of agent-based modeling. Starting from theoretical hypotheses about generative mechanisms, we tried to generate generic and specific stylised facts, with a model as parsimonious as possible. The progressive evaluation of increasingly complex models led to the satisfactory simulation of observed urban evolution and highlighted specific trajectories that “resist” modeling.

Key-words. System of cities, Post-Soviet Space, Soviet Union, Generic/Specific, Simulation, Urban evolution, Urban growth.

Link. HAL-SHS.

PhD Clara Schmitt

Schmitt C., 2014, Modeling settlements systems dynamics : from SimpopLocal to SimpopNet, University Paris I.

Abstract. Is urban growth the result of multiple interactions between cities ? Urban evolutionary theory (Pumain, 2000), based on this postulate, analyses urban growth processes. This thesis, undertaken in an inter-disciplinary context, aims to evaluate the validity of the hypothesis by means of computer simulation. Strong regularities of the urban systems dynamics are extracted from the accumulated scientific knowledge and synthetized into ten major stylized facts. Two simulation models, SimpopLocal and SimpopNet, are then built, documented – thanks to a standardized grid – and systematically explored. They each question a specific aspect of the urban evolutionary theory : the nature of the inter-urban interactions for the first model (i.e. competition for innovation) and their support for the second model (i.e. the role of the communication network structure). The evaluation of the two models required the design and the implementation of two original exploration protocols : an automated calibration method and a sensibility analysis protocol (the Exploration Profile algorithm) which individually evaluates the contribution of each implemented mechanism to the simulated behavior. These two forms of exploration systematically confront the simulation results with current scientific knowledge. They indicate that the two models are able to account for key processes of urban systems dynamics, such as their hierarchical organization, and demonstrate for the first time the need for interurban interaction mechanisms in order to simulate urban evolutions that are close to those observed on real urban systems.

Key-words. System of cities, Urban systems, Simpop, SimpopLocal, SimpopNet, Simulation, Urban evolution, Urban growth.

Link. HAL-SHS.

Phd Ignazzi

Ignazzi C. A., 2015, Coevolution in the Brazilian Urban System, University Paris I.

Abstract. This thesis analyzes the urban system in Brazil adopting an advanced database that have been constructed collecting demographic data in order to examine the evolution of the population of all Brazilian agglomerations since the first Brazilian official census carried out in 1872 until 2010.
The largest country of South America has already completed its urban transition during the last century and is characterized by the contrast between a larger number of small towns throughout the immense territory and enormous Metropolitan areas dominating the system of cities.
Despite its geographical and historical peculiarities, this system shares with others in the world the same properties of hierarchical differentiation and urban growth processes (Zipf’s law and Gibrat’s model).
Economic data have been integrated in the database with the aim of testing the validity of scaling laws for Brazil and performing robust statistical analysis in order to explore the functional differentiation of cities, their economic performances and the spatial autocorrelation processes occurring among them.
The most interesting result is characterizing the Brazilian urban hierarchy over the long period and measuring the increasing inequalities in city sizes. Moreover, the parallel support of demographic and economic data is essential to identify the connection between population and economic growth in one of the most urbanized country of the world.

Key-words. System of cities, Urban systems, Urban hierarchy, Zipf, Gibrat, Urban trajectories, Brazil, Functional differentiation, spatial autocorrelation.

Link. Online.

PhD Rey-Coyrehourcq

Rey-Coyrehourcq S., 2015, An integrated platform for building and evaluating model of simulation in geography, University Paris I.

Abstract. Since 1990’s, Agent Based Modelling are commonly used by geographers to study complex systems like cities.
However, very few technical platforms are advanced by searchers to assist in the construction and evaluation of models of simulation. With the help of ERC program GeoDiverCity and the formation of an expert interdisciplinary team, we try to solve these problematic following two objectives. Relying on the support of OpenMOLE platform developed at the Institute of Complex System Paris-Ile-de-France in order to make it simple the distribution of simulation on distributed computing environments, we identify, use or build new tools and methodology to construct and explore model of simulation. To anchor this work in practice, we use this platform to build and explore a new model of simulation: SimpopLocal. This very practical work is accompanied by an historical and epistemological reading of simulation, and the means of simulation in geography. These contextualisation permits us to examine, and perhaps to anticipate, the historical link between the old problematic of “Validation”, very important to consider for knowledge justification, and the building and exploration of models of simulation.

Key-words. Simulation, Agent Based Modelling, Complex systems, Distributed computing environments, Validation.

Link. Online.

PhD Solène Baffi

Baffi S., 2015, Railways and city in territorialization processes in South Africa : from separation to integration?, University Paris I.

Abstract. First railway network in Africa, the South African Railways constitute a privileged marker of the territorial mutations that have been shaping this country for decades. The radicalism of political systems shows through the persistency of segregative schemes of which the railways, as part of the planning toolbox, are one of the elements. The inertia of this infrastructure questions its re-appropriation and insertion into the various planning projects over the long term. This thesis approaches the long-term dynamics of ‘territorialisation’ in South Africa through the prism of railways. This study focuses on the interaction between cities and the railway network, at both the interurban and the intra-urban levels.
In order to understand this interaction, cities’ location and railways diffusion patterns are analysed, with a specific emphasis on the shape of the network. Indeed, through their pattern, railways express the intentions of actors in charge of planning at the national level. By the flows it supports and the mobilities it enhances, it gives us information on the socioeconomic requirements of society and the power relations it contains. Thus, this thesis relies on a qualitative and quantitative approach aiming to outline the structuring effects of the railways in South Africa over the long term.
Its political use by successive segregationist powers explains partly why nowadays railways keep on marking partition in the post-apartheid urban space and in the practices of city dwellers. However, the recent rail renewal observed in the metropolises, Cape Town in particular, might indicate a possible major inflection in the persistency of inherited dynamics. Indeed, the evolution of the rail offer demonstrates a shift towards a co-construction trend between South African society and urban planning unheard of until now.

Key-words. System of cities, Urban system, South Africa, Railways, Territorialisation, Inclusion/exclusion, Intra-urban mobilities.

Link. HAL-SHS.

PhD Olivier Finance

Finance O., 2016, French cities hosting foreign direct investments: from networked companies to localized establishments, University Paris I.

Abstract. Transnational corporations, which are amongst the major players in the contemporary global economy, integrate and exclude territories at various scales, due to their specific location strategies. These inequalities are well known at an international scale, yet the knowledge of this diverse integration is much more limited regarding urban levels, although cities and metropolises are considered as being the major nodes of the globalized networks. France and the OECD countries certainly appear in central positions in the networks that characterize these corporations, but observations made at the urban level remain very partial due to the lack of localized data. This thesis suggests to both approach and localize conventional data about Foreign Direct Investment (FDI) in the French case by mobilizing data about financial links connecting economic units. The detailed breakdown of transnational corporations affiliation networks has been conducted up to the level of the establishments, which are the real individual economic and geographic cells of these transnational networks. An original database about localized inward FDI stocks has been built and explored to appreciate how far foreign transnational corporations integrate the 355 main cities into the French urban system. These data revealed the diverse integration of French cities, between dependence and attractiveness for the investors. The mobilization of scaling laws, which constituted a major analytical tool in this work, allowed us to identify the major factors explaining the diverse integration of French cities into the whole system of cities, reflected both by inequalities of hierarchical and regional order.

Key-words. System of cities, Urban system, France, Scaling laws, Foreign Direct Investment, Transnational firms.

Link. HAL-SHS.

How to estimate the singularity of (post-)Soviet cities?

Revue d'Economie Regionale et Urbaine - n°1 - 2017 New Article, in RERU – Revue d’Économie Régionale et Urbaine.

Cottineau C., 2017, “How to estimate the singularity of (post-)Soviet cities?”, Revue d’Économie Régionale et Urbaine, n°1-2017.

Abstract. Although the urbanisation of the Former Soviet Union is a unique experience, this article questions the possibility of estimating the share of the generic processes of urban growth, spatial location and economic specialisation, the share of the particular processes of urbanisation linked to the size and political organisation of the country, and the share of the singular processes that cannot be explained without a local knowledge of the events which happened in the cities under study. Using several types of models at different scales, we identify the residual as the element that “resists modelling” and illustrates the singular evolutions of the Soviet Union and its cities. To do so, we built a harmonised urban database and fitted hierarchical, spatial and regression models. We conclude that city size inequality increased in a generic manner compared to other systems of cities, that the spatial distribution of cities resemble that of vast countries (especially with the increased reliance on sub-surface resources), and that there exists a set of singular urban trajectories.

Key-words. System of cities, Soviet Union, Singularity, Models, Urban growth, Urban trajectories, Urbanization.

Link. Revue d’Économie Régionale et Urbaine

The Brazilian Urban System: the trajectories of Brazilian cities between general dynamics and specific peculiarities

Brazil-citiesNew Article, in Cybergeo, European Journal of Geography.

Ignazzi C. A., 2015, «The Brazilian Urban System: the trajectories of Brazilian cities between general dynamics and specific peculiarities», Cybergeo : European Journal of Geography, Systems, Modelling, Geostatistics, document 754, DOI:10.4000/cybergeo.27349

Abstract. This article analyzes the dynamic of Brazilian system of cities illustrating the growth of Brazilian population through its own historical peculiarities like internal and international immigration, industrialization, metropolization. Urban hierarchy is revealed by means of the Zipf’s law and Gibrat’s model is used to describe the mechanisms of urban growth. Deviations from this model were explained by studying the specific trajectories of five different classes of cities, on various time scales (from 1872 to 2010 and from 1960 to 2010). These deviations highlight different kinds of factors (political, economic, localization of resources and historical accidents) that shaped the Brazilian system during the last century.

Key-words. Spatial autocorrelation, Urbanization, Brazil, Urban Hierarchy, Trajectories, Zipf’s law, Gibrat’s law, Markov Chains.

Multilevel comparison of large urban systems

New publication, in Cybergeo, European Journal of Geography

Pumain D., Swerts E., Cottineau C., Vacchiani-Marcuzzo C., Ignazzi C.A., Bretagnolle A.,  Delisle F., Cura R., Lizzi L., Baffi S., 2015, « Multilevel comparison of large urban systems », Cybergeo : European Journal of Geography [En ligne], Systèmes, Modélisation, Géostatistiques, document 706, URL : ; DOI : 10.4000/cybergeo.26730

Abstract. For the first time the systems of cities in seven countries or regions among the largest in the world are made comparable through the building of spatio-temporally standardised statistical databases. We first explain the concept of a generic evolutionary urban unit (“city”) and its necessary adaptations to the information provided by each national statistical system. Second, the hierarchical structure and the urban growth process are compared at macro-scale for the seven countries with reference to Zipf’s and Gibrat’s model: in agreement with an evolutionary theory of urban systems, large similarities shape the hierarchical structure and growth processes in BRICS countries as well as in Europe and United States, despite their positions at different stages in the urban transition that explain some structural peculiarities. Third, the individual trajectories of some10,000 cities are mapped at micro-scale following a cluster analysis of their evolution over the last fifty years. A few common principles extracted from the evolutionary theory of urban systems can explain the diversity of these trajectories, including a specific pattern in their geographical repartition in the Chinese case. We conclude that the observations at macro-level when summarized as stylised facts can help in designing simulation models of urban systems whereas the urban trajectories identified at mico-level are consistent enough for constituting the basis of plausible future population projections.

Key-words. Urban systems, Zipf, Gibrat, Cities trajectories, BRICS

Selective investments of foreign transnational firms in the French system of cities revealed by scaling laws

Many economic links connect the cities embedded in systems of cities (for example commercial links, financial ownership links…). They participate in the intrinsic complexity of these systems. The interactions and connections between places due to the economic stakeholders can substantially impact the shape and the dynamics of any system of cities. Especially, networks built up by transnational firms by the ownership of subsidiaries located beyond their territorial borders into foreign systems of cities, could shape the future of related cities. These foreign investors can provide these cities new jobs by the creation or extension of establishments but also sometimes weaken them through a massive control on their total employment, transnational investments being expected to be more volatile than national investments. A new source of data providing interesting insights on the effective location of these impacts on French cities is now available (see box below).

An original database

Three data bases were combined to assign the establishments (physical location of production, selling points) controlled by foreign capital in French cities (=“aires urbaines” i.e. SMAs):

  • the ORBIS data base (produced by Bureau van Dijk, augmented by Céline Rozenblat – UNIL – GeoDiverCity) that contains all financial linkages between companies into the 3,000 major groups worldwide,

  • the LIFI data base (produced by INSEE), a French data base similar to the previous one yet extended to smaller groups,

  • the CLAP data base (also produced by INSEE), to add the level of establishments (the most suitable one for our geographical approach, whereas ORBIS and LIFI are limited to companies).


A strong hierarchical effect…

Our first analyses outline the high concentration of foreign-controlled employment in the biggest cities of the system (figure 1). We could expect this result by observing these networks at the level of the companies, as it is known that their headquarters are more concentrated than their physical locations. Yet even when these jobs are located at the level of establishments, the foreign-controlled activities scale superlinearly (as indicated by the scaling law exponent higher than 1). Foreign-controlled employment is consequently much more concentrated in largest cities. The distribution of these establishments is not proportional to the size of the cities, as it benefits first and foremost to the biggest cities of the system.

Figure 1: Important role of the urban hierarchy in the distribution of foreign-controlled jobs in the French system of cities


and regional effects

Taking into account the power law relationship and the quality of fit which remains nevertheless medium, another major trend appear in our first analysis. The amount of foreign-controlled jobs in the smallest cities is highly variable, some of them being deeply invested by the transnational firms while some others are almost avoided. A more detailed study of this variability show strong regional effects in addition to the hierarchical ones (figure 2). Cities with higher amounts of foreign-controlled jobs than expected by the scaling law are almost always located in Northern and Eastern parts of the country or located near Paris; cities where this amount is lower than expected are located in Southern and Western parts of the country. Thus foreign investors adapt to former spatial trends of urban functional specialization in France.

Figure 2: Regional effects in the distribution of foreign-controlled jobs within the smallest cities of the system


Further analyses

As urban hierarchy doesn’t explain the distribution on its own, further analyses will investigate the effects of the economic specialization of cities, the closeness to bigger cities, or the shape of the networks where these cities are embedded or not. Our intention is to formulate some stylized facts identifying emerging properties and network dynamics that characterize the distribution of the foreign transnational firms into the French system of cities at an unprecedented level of detail. Besides classical multivariate analysis and networks analysis, scaling laws will constitute one of our main analysis tools, as they shed new light on the connection between urban functions, city sizes and economic innovation cycles (Pumain et al., 2006). This thesis, part of the GeoDiverCity program, will contribute an approach historicizing the scaling laws parameters relating them to the innovation cycles and the hierarchical diffusion of innovations theory in the system of cities.

Olivier Finance, PhD student under the supervision of Denise Pumain and Fabien Paulus.

GeoDiverCity at AAG Annual Meeting 2014 in Tampa, Florida

Various contributions of GeoDiverCity team are scheduled for the Annual Meeting of the Association of American Geographers in Tampa, Florida :

Tuesday, 4/8/2014

> Session : 1654 European Research Council – Top European grants for brilliant minds from across the world, from 4:40 PM – 6:20 PM in Grand Salon C, Marriott, Second Floor. Organizer : Katja Meinke.

17:15-17:30    Denise Pumain, “ERC from an Advanced Grantee’s perspective.”

Wednesday, 4/9/2014

> Session : 2268 Urban systems and scaling laws: Functional diversity and urban economic trajectories, from 10:00 AM – 11:40 AM in Meeting Room 1, Marriott, Second Floor. Organizer : Céline Vacchiani-Marcuzzo.

10:00-10:20    Elfie Swerts, Céline Vacchiani-Marcuzzo, Fabien Paulus, “Scaling laws as a tool for characterising the functional evolution in urban systems”

10:20-10:40    Olivier Finance, “Transnational firms in the French system of cities and scaling laws”

> Session : 2239 Geosimulation Models 1: Methodological Advances, from 10:00 AM – 11:40 AM in Room 39, TCC, Fourth Floor. Organizers : Paul Torrens, Suzana Dragicevic, Andrew Crooks.

11:20-11:40    Mathieu Leclaire, Romain Reuillon, “Simpuzzle/Janet tools or how to build a step by step modular ABM ?”

> Session : 2539 Geosimulation Models 3 : Applications – Macro, from 2:40 PM – 4:20 PM in Room 39, TCC, Fourth Floor. Organizers : Paul Torrens, Suzana Dragicevic, Andrew Crooks.

14:00-14:20    Clémentine Cottineau, Paul Chapron, “Evaluation & Calibration for the comparison of ABMs of cities’ trajectories”

16:00-16:20  Denise Pumain, Clara Schmitt, Sébastien Rey-Coyrehourcq, Romain Reuillon, “Building and exploring an agent-based model with OpenMOLE”

The Integration of the cities in the networks of multi-national firms in the agri-business industry

The development of the agrofood sector has always taken place alongside the process of urbanisation (Bairoch, 1988).  The agrofood sector, today the primary global manufacturing sector, offers evidence of former processes of integration of cities undergoing globalisation.  Each city, according to its size, its attractivity, its power in the network, its capacity or lack of capacity to connect with the economic actors, occupies a particular position in these agrofood companies’ networks

Starting from original information about the networks of the subsidiaries of the largest agrofood companies, we developed indicators describing the weight and the centrality of the cities in these networks . Data about the financial linkages between agrofood firms deriving from the Orbis base, 2010, that we prepared (Orbis, Bureau Van Dijk, 2010 ; Rozenblat, 2010).  To highlight the position of each city in the global strategies of agrofood companies we carried out a principal component analysis and an ascendant hierarchical classification according to six variables:

  • The population;
  • The number of subsidiaries present in each city;
  • The intra-urban connectivity corresponding to the mean number of the relations of a subsidiary in a city;
  • The centrality of intermediarity (betweenness centrality)
  • The degree (total number of relations), In-Degree (relations entering) and Out-Degree (relations going out);

An indicator of power as being the difference between relations going out and relations entering, relativised by the total number of relations.

The principal component analysis brings out two main dimensions, which focus on 80% of the information contained in the data.  The first factor, representing 62.5% of the information, distinguishes between the cities according to their attractivity and their local and global centrality in the network (inter-urban dimension).  The second factor summarizes 16.5% of the information, opposing small and middle-sized cities benefiting from a strength associated with their strong intra-urban connectivity to cities with a large population but less attractive (intra-urban dimension).

The ascendant hierarchical classification identifies 6 classes of cities, and a Chi2 test confirms a significant relationship between the continental membership of the city and its classification.

  • Class 1 is represented by small to middle-sized cities with weak centrality and weak attractivity This class brings together 70% of the cities in our sample.  All the continents are represented in it in a homogeneous fashion.
  • Class 2 is made up of cities with a strong population and weak scores for attractivity and centrality in the network. It brings together 33 cities ; the Asian cities are over-represented in this class :  they represent 2/3 of the  group, followed by the cities of South America, also over-represented (7 cities).
  • Class 3 describes small to middle-sized cities, controlling (strong indications of power) with a strong intra-urban connectivity. The European cities are largely over-represented, as they represent 70% of the group (119 cities). The Asian and South American cities are under-represented.
  • Class 4 represents middle-sized cities with strong indications of centrality. Once again the European cities are over-represented, they represent 70% of the group.
  • Class 5 defines cities with very strong centrality indications of degree and of ‘intermediarity’ (Betweenness). 12 cities belong in this class:  5 European cities, 3 Asian cities and 3 North American cities; 1 African city (Johannesburg).
  • Class 6 describes the cities of Paris and London, which have an exceptional position in the network, with a very strong betweenness centrality, and of relatively weak ‘in-and-out’ degrees.

These results demonstrate that the cities do not have the same attractivity in relation to their population size.  With equal populations, the parameters of the power functions of the scaling laws of the systems of the North American and European cities are twice as strong as those of the system of Asian cities.  These relationships demonstrate that the economies of agglomerations are superior in the North American and European cities, which is probably linked to the quality of their infrastructures, to the diversity of their economic actors, and to their position in the global value chains.

This typology also brings to light a strong centre-to-periphery structure, with at its head, London, the cradle of the food-processing industry, the most attractive and central city; then Paris, less attractive than London, but which plays the particular role of international bridge in the agrofood companies’ networks. In the second position are found some international cities that are integrated and central, but whose influence varies from the intra-continental scale to the national scale.  The periphery is defined by the cities of Groups 1 and 2.  Not surprisingly, the cities of this group are for the most part located in poor countries : most of them being Asian cities, cities of Africa, South America, and the Pacific coast of the North American continent.

Bérengère Gautier

BAIROCH P. (1988) Cities and economic development: From the dawn of history to the present, University of Chicago Press, 596 pages

GAUTIER B. (2012) « Intégration et développement des villes méditerranéennes par les réseaux de firmes multinationales du secteur agroalimentaire », Université de Lausanne, Thèse de doctorat, 322 pages.

ROZENBLAT C. (2010) “Opening the Black Box of Agglomeration Economies for Measuring Cities’ Competitiveness through International Firm Networks”, Urban Studies, Vol 47, n°13, pp  2841-2865

The dispersion of added value in Russia through multinational networks

This map is made out from an analysis of the multinational firms’ networks in Russian cities (ORBIS database, Bureau van Dijk, 2010; C. Rozenblat). The links between owners and subsidiaries are aggregated into urban agglomerations, and differentiated by activity sectors (NACE). Company groups working in Russia have been constituted : chains of ownership were formed where subsidiaries are owned with a share of at least 50%. The map shows mean points and standard distances of head groups’ locations in Russia. The more we travel East from Moscow, the less the share of added value in production: The barycentre for finance, information and communication activities (that is : advanced or metropolitan services) is the most western one, the closest to Moscow, and their standard distance is 2000km;  further to the East is the barycentre of trade groups, similarly scattered, followed by manufacturing industries, that are less concentrated; eventually, mining and transport groups have a remote gravity centre located at thousand kilometres to the East, with the largest spatial dispersion.

Cyril Jayet and Clémentine Cottineau